City University of New York (CUNY) City University of New York (CUNY)
CUNY Academic Works CUNY Academic Works
Dissertations and Theses City College of New York
2019
Void fraction measurements in a large (0.3 m) diameter bubble Void fraction measurements in a large (0.3 m) diameter bubble
column using Wire Mesh Sensors and Pressure Transducers: column using Wire Mesh Sensors and Pressure Transducers:
Effect of Spargers Effect of Spargers
Afolabi G. Ojo CUNY City College
How does access to this work benefit you? Let us know!
More information about this work at: https://academicworks.cuny.edu/cc_etds_theses/878
Discover additional works at: https://academicworks.cuny.edu
This work is made publicly available by the City University of New York (CUNY). Contact: [email protected]
1
Void fraction measurements in a large (0.3 m) diameter bubble column using
Wire Mesh Sensors and Pressure transducers: Effect of spargers
THESIS
Submitted in partial fulfillment ofHom
the requirement for the degree
Master of Engineering (Mechanical)
at
The City College of New York
of the
City University of New York
by
Ojo Afolabi Gabriel
May 2019
Approved:
Professor Masahiro Kawaji, Thesis Advisor
CUNY Energy Institute
Professor Feridun Delale, Chairman
Department of Mechanical Engineering
2
Abstract
The importance of the accurate temporal and spatial measurements of the two-phase flow
parameters in bubble columns is very well known. The aim of this research work is to report the
spatial measurements of the void fraction distribution in a 30 cm diameter cylindrical bubble
column using Wire Mesh Sensor (WMS) tomography and Pressure Transducers. Pair of WMS
sensors, with a 64×64 wire configuration of each sensor, were installed which are separated by a
distance along the axis of the bubble column. Wire Mesh Sensors and PTs Data were collected for
time-averaged and transient with a sampling frequency of 1000 Hz respectively. The principle of
Wire-Mesh Sensors (WMS) and algorithms for estimating the void fraction from the WMS raw
experimental data have been discussed in great detail. The void fraction results obtained from the
WMS are compared against the void fraction results obtained from the Pressure Transducers.
Experiments were performed for seven superficial gas velocities, to cover both Homogenous and
Heterogenous flows. Two different spargers (point sparger and four arm spargers) were used to
study the effect of the sparger design on the hydrodynamics. The measurement uncertainty of the
WMS and PTs for air-water two-phase flows is investigated by repeating the experiments. The
effect of superficial gas velocity and sparger design on the radial and steady-state void fraction
profiles have been reported. These experimental investigations disclose the quantitative
measurements of the steady-state flow parameters which have not so far been reported, and this
data can also be used by the theoretical CFD modelers for their simulation data validation.
3
Acknowledgment
This thesis becomes a reality with the kind support and help of many individuals, I would like to
extend my sincere thanks to all of them.
Foremost, I want to offer this endeavor to our GOD Almighty for the wisdom he bestowed upon
me, the strength, peace of mind and good health in order to finish this research.
I would like to express my gratitude towards my family for the encouragement which helped me
in completion of this paper.
I am highly indebted to the Grove School of Engineering and CUNY Energy Institute for the
platform provided to pursue this research.
I would like to express my special gratitude and thanks to my adviser, Professor Masahiro Kawaji
for imparting his knowledge and expertise in this study.
My instructor, Dr. Dinesh Kalaga for sharing his knowledge and technical know-how.
My thanks and appreciation also go to my colleague, Heber Blanco De Brand and people who have
willingly helped me out with their abilities.
4
Table of Contents
1. Introduction ........................................................................................................................................... 5
2. Experimental Test Facility .................................................................................................................... 7
3. Experimental Methods and Data analysis ........................................................................................... 13
3.1. Expanded height method .................................................................................................................. 13
3.2. Pressure Drop Method ..................................................................................................................... 13
3.3. Wire Mesh Sensor ............................................................................................................................ 14
3.3.1. Void fraction analysis for wire mesh sensor data. .................................................................... 16
4. Results and discussion ........................................................................................................................ 19
4.1 Overall volume averaged hold-up ..................................................................................................... 19
4.2 Effect of Gas velocity ....................................................................................................................... 22
4.2.1 Near the gas sparger ....................................................................................................................... 22
4.2.2 In the fully developed region ......................................................................................................... 26
4.3 Effect of initial liquid height ............................................................................................................. 32
4.4. Effect of sparger design ................................................................................................................... 36
5. Conclusions ......................................................................................................................................... 38
6. An additional chapter on the Very-High-Temperature Reactor (DOE Project) and the Drilling Fluid
(PIRE Project): ............................................................................................................................................ 39
References: .................................................................................................................................................. 41
5
1. Introduction
Several important industrial processes require a reaction between a gas and a liquid to
produce a particular product, or to absorb a gaseous component in a liquid. According to Tatterson
(1991), 25% of all chemical reactions occur between a gas and a liquid. In some of these processes,
a large liquid bulk is required because the chemical reaction between the dissolved component and
a component of the liquid phase is slow with respect to gas-liquid mass transfer, limited usually
by the liquid side. In these cases, a gas-liquid bubble column (a vessel filled with liquid equipped
with a sparger at the base for dispersing gas) is a suitable reactor, which is therefore extensively
used in the process industries (often preferred to agitated tanks because bubble columns do not
have any moving mechanical parts). Gas-liquid bubble columns offer some distinct advantages:
among which their relatively simple construction, low operating costs, excellent heat transfer
characteristics to immersed surfaces and the ease with which the liquid residence time (in
continuous liquid operations) or circulation time (in batch liquid operations) can be varied, should
be particularly noted (Shah et al. (1982); Fan (1989)). However, many important fluid dynamical
aspects of the gas-liquid two-phase flow prevailing in bubble columns are still poorly understood
and difficult to predict a priori. This unsatisfactory state of the art has led to an increased interest
in recent years in detailed modeling of bubble column (Krishna et al. (1991); Sokolichin et al.
(1997); Buwa and Ranade (2002); Sanyal et al. (1999)) and in the development of advanced
experimental tools (Devanathan (1991); Chen and Fan (1992); Krishna et al. (1994); Degaleesan
(1997); Pfleger et al. (1999) etc.). In spite of these very significant efforts, the design of bubble
columns is still difficult due to the complex hydrodynamics and the inherent unsteadiness of the
liquid flow generated by the passage of the bubbles. The correct and prudent design of bubble
6
columns requires the accurate identification of all the relevant flow patterns and holdup profiles
under different experimental conditions, and their measurement, which is key to the selection of
appropriate models for simulating the two-phase flows. Further, Due to their low operating cost
and simple construction, bubble columns are among the most omnipresent gas-liquid multiphase
contactors in chemical, petrochemical and biochemical industries. The flow patterns and the flow
parameters like void fraction distribution, bubble size distributions, liquid phase turbulence are
interrelated with operating and design variables in an untidy manner; hence the understanding of
the hydrodynamics is very important. The void fraction is an important design parameter which
governs the performance of bubble column reactors. It is well-known that the void fraction depends
on the mean bubble size, number of bubbles and bubble velocity besides, dependence operation
on operating and design variables. For the efficient and reliable design of bubble column reactors,
the development of fluid dynamic models is indispensable to understand these complex
interactions, which requires temporally and spatially resolved axial and radial void fraction
profiles.
Over the past few decades, several invasive and non-invasive measurement techniques to
measure the void fraction and mean bubble size in gas-liquid flows have become available. Some
examples include optical probes for transient/steady local void fraction measurement, Electrical
Resistive Tomography (ERT) or Electrical Capacitance Tomography (ECT) for transient/steady
void fraction measurement over a cross section, Single-beam or Multi-beam γ-ray Densitometry
for steady state void fraction measurement over a chord or plane. Each of this measurement
technique has its limitations and needs to be evaluated for each application. Wire Mesh Sensor
(WMS) is an invasive measurement technique, which offers the transient and steady void fraction
data measured over a plane in both batch and continuous gas-liquid flows.
7
In the last several decades, many researchers have reported the experimental investigations
with more focus on the steady state measurements. There is little information available on the
context of “development of flow”, in other words, the transient behavior of two-phase flow at gas
injection. Therefore, the present paper makes an attempt to gain insights into the fundamentals of
hydrodynamics by measuring the temporally and spatially resolved void fraction data in a 30 cm
diameter bubble column. The high data acquisition rate and spatial resolution give information
about the flow in greater detail. Further, the data analysis such as radial void fraction, cross-
sectional void fraction distribution is presented and discussed.
2. Experimental Test Facility
The schematic of the cylindrical bubble column with the location of pressure transducers
and wire mesh sensors is given in Figure 1A. The column is made of Perspex having an internal
diameter of 30cm and height of 2m. Experiments were carried out with air-deionized (DI) water,
air-tap water, air - 20 PPM Na2CO3 dissolved in DI water and air- 100 PPM CaCl2 dissolved in
DI water systems. Necessary care has been taken by adding a few ppm of salt to maintain the
minimum liquid conductivity needed for the use of wire mesh sensors. The compressed air was
passed through a series of filters, to remove the dust and oil particles and the gas flow rate was
measured using a calibrated rotameter. The initial liquid height was varied by three initial levels
for all the experiments (Level-1: 100 cm, Level-2: 142 cm and Level-3: 183) and two spargers
(one with point sparger and the other was four arm sparger) with 5 holes of 1mm diameter on each
arm (Figure 1B) of the used four arm gas sparger. Two wire mesh sensors (Figure 1C) were placed
at four different axial locations to capture the axial variation of the gas hold up distribution along
the column height (details can be found in Figure 2). Six pressure transducers (Omega: PX-309)
were placed strategically to measure the volume averaged gas fraction along the height of the
8
reactor. The air volumetric flow rate is varied between 5 CFM to 35 CFM which corresponds to
the range of air superficial gas velocity from 0.037 mm/s to 0.26 mm/s. Further, the overall volume
averaged gas hold-up was calculated by measuring the initial and expanded liquid heights for each
gas velocity. Each experiment was repeated three times to check the reproducibility and reliability
of measurement techniques.
Figure 1. (A) Schematic of the experimental set-up, (B) Four arm sparger, (C) Photograph of the
wire mesh sensor
9
Figure 2A: Front and isometric views depicting the initial water levels and the location of
wire mesh sensor for the Configuration: 1 (dimensions are in cm)
10
Figure 2B: Front and isometric views depicting the initial water levels and the location of
wire mesh sensor for the Configuration: 2 (dimensions are in cm)
11
Figure 2C: Front and isometric views depicting the initial water levels and the location of
wire mesh sensors for the Configuration: 3 (dimensions are in cm)
12
Figure 2D: Front and isometric views depicting the initial water levels and the location of wire
mesh sensors for the Configuration: 4 (dimensions are in cm)
13
3. Experimental Methods and Data analysis
3.1. Expanded height method
Before starting the experiment, care has been taken to remove the air trapped in the gas injection
line. The overall gas hold-up was calculated by measuring the static bed height (HS) and
expanded/dynamic bed (HD) heights. A measuring scale is attached to the column to avoid the
errors associated with measuring the fluctuating (expanded) bed height. The overall gas hold-up (
G ) was estimated using the following relation (kalaga et al. 2017).
D SG
D
H -H=
H (1)
3.2. Pressure Drop Method
The volume averaged void fraction data were obtained using the pressure drop method, which
entails using two pairs of pressure transducers placed close to each WMS plane (Figure 1). The
pressure drop method is based on the underlying principle for the volume average void fraction,
or gas hold up, which can be derived from the Bernoulli equation in a steady fluid along a pipeline:
1
2ρv2 + ρgh + P = constant (2)
where 1
2ρv2 is kinetic energy, ρgh is potential energy and P is pressure. Considering the two points
where the pressure transducers are located, Eq. (2) can be expressed as:
1
2ρmv2
1 + ρgh1 + P1 =1
2ρmv2
2 + ρgh2 + P2 (3)
where ρm is gas-liquid mixture density, defined as:
ρm = (1 − α)ρL + αρG (4)
14
Since the pipe has a uniform cross-sectional area, it is assumed force due to the velocity of the
fluid is constant, v1 = v2 = v. Also, the location of the pressure transducer 1 (H/D = 1.4) is
considered as a reference point (h1 = 0) and the height of pressure transducer 2 is considered to be
h2 = h.
Substituting Eq. (4) into Eq. (3) and simplifying (Hernandez et al. 2017);
∆P = [(1 −∈L)ρL +∈G ρG]gh (5)
The void fraction can be obtained from Eq. (5) gives
∈L=ρL−
∆P
gh
ρL−ρG (6)
Next, the volumetric average void fraction calculated from the pressure transducers has the offset
and to account the pressure transducer’s offset the following equation is used to calculate the void
fraction (Mutharasu et al. (2018))
𝜖�̅�,12 =(𝑃2−𝑃1)𝑇𝑃−(𝑃2−𝑃1)𝐿
(𝑃2−𝑃1)𝐿−(𝑃2−𝑃1)𝐺 (7)
The volume averaged void fraction between two pressure transducers P1 and P2, Where,
(𝑃2 − 𝑃1)𝐺 - pressure drop recorded for gas.
(𝑃2 − 𝑃1)𝐿 - pressure drop recorded for liquid.
(𝑃2 − 𝑃1)𝑇𝑃 - pressure drop recorded for two-phase gas-liquid dispersion.
3.3. Wire Mesh Sensor
Wire mesh sensors found in the literature can be classified into two types; capacitance wire
mesh sensor (Silva, 2008) which is based on the permittivity of the fluids and conductance wire
mesh sensor (Prasser et al., 1998) which is based on the conductivity of the fluid. The working
15
principle of the capacitance WMS is to calculate the local instantaneous void fraction data from
the measured local values of the fluid conductivity. A pair of capacitance-based wire mesh sensors
with a maximum temporal resolution of 10 KHz and spatial resolution of 4 mm ×4 mm have been
used to quantify the flow parameters. Wire mesh sensor is characterized by a matrix-like
arrangement of crossing points formed by two parallel planes positioned orthogonally and
separated by a vertical distance of 3.15 mm. Each sensor has two planes, which also have 64
stainless steel electrode wires (diameter=0.4 mm) uniformly distributed over the circular cross-
section for measuring the conductivity of the flowing mixture around each one of the 64 nodes.
One wire plane of the sensor is used as a transmitter and another wire plane is used as a receiver
(Figure 1). Each transmitter wire is activated sequentially by the voltage pulses supplied by the
electronics, while all other transmitter wires are kept to ground potential, the current received by
the receiver wires at each crossing-point node is then recorded by the data acquisition system. The
voltage recorded by a node is high when surrounded by the water and voltage is low when
surrounded by air. Further details of the WMSs operation and principle can be found in Prasser et
al. (1998). For the experiments, a pair of WMSs were placed at different elevations from the
bottom of the reactor. The sensor was sandwiched between PVC flanges which allow the fixation
of the sensor into the reactor. Wire mesh sensor data was recorded for the duration of 240 secs at
a frequency of 1000Hz for steady-state measurements and transient measurements.
16
Figure 3A: Schematic representation wire mesh sensor configuration and the associated
electronics.
3.3.1. Void fraction analysis for wire mesh sensor data.
In order to calculate the void fraction distribution, the voltage data obtained from the WMS
needs voltage values of each node generated from a chosen calibration method. There are two
possible different methods of calibration namely, histogram calibration method and water
calibration method. Histogram calibration method requires the histogram of the frequency of
occurrences of the voltage signal for both the phases at each node of the sensor. This histogram
data usually has two maxima, one at low voltage for the gas phase and other at high voltage for
17
the liquid phase. This method does not require separate calibration data as the same measurement
data can be used to extract the calibration values. This method of calibration cannot be used for
two-phase flows in horizontal channels and the churn turbulent flows wherein very few sensor
nodes are available for pure water. For water calibration method, the sensor nodes should be
completely immersed in water. This method of calibration offers an advantage owing to the fact
that this data can be applied for all gas velocities, but proper care should be paid to the fact that
the operating fluid’s temperature and conductivity need to be the same. In the present work, water
calibration method is used, and the calibration data is obtained by filling the column with water.
The instantaneous local void fraction at each node is obtained by the following equation:
∈𝑖,𝑗=𝑆_𝑊𝑖,𝑗−𝑆_𝑇𝑃𝑖,𝑗
𝑆_𝑊𝑖,𝑗 (1)
The time-averaged void fraction distribution is computed using the following equation:
∈̅𝑖,𝑗=1
𝑘𝑚𝑎𝑥∑ ∈𝑖,𝑗,𝑘
𝑘𝑚𝑎𝑥𝑘=1 (2)
The variation of the area-averaged void fraction over the measurement time period (transient void
fraction ∈ (𝑡) profile) is obtained using equation (2). The area averaging is based on the weight
coefficient, which defines the contribution of each crossing node in the sensor matrix to the column
diameter. For example, if the mesh node is inside the vessel ai,j×Asensor=∆X×∆Y (A in Figure 3)
and If the mesh node near the column wall ai,j×Asensor<∆X×∆Y (B in Figure 3).
∈̅𝑘=∈ (𝑡) = ∑ ∑ 𝑎𝑖,𝑗 .𝑗 ∈𝑖,𝑗,𝑘𝑖 (3)
The time-averaged and area-averaged void fraction can be obtained by combining the equations 2
and 3. which is given by Eq. 4.
∈̅= ∑ ∑ 𝑎𝑖,𝑗.𝑗 ∈𝑖,𝑗,=𝑖1
𝑘𝑚𝑎𝑥∑ ∈̅𝑘
𝑘𝑚𝑎𝑥𝑘=1 (4)
18
For the radial void fraction profiles, the sensor geometry is divided into hypothetical ring-shaped
domains (m=80 in the present study). Importantly, the 2-D void fraction distribution data from the
WMS was averaged over 80 rings of different radii, r, and r+dr. Eq. 5 is used to calculate the
azimuthally averaged void fraction data for a given measurement period.
∈̅𝑚=1
𝑘𝑚𝑎𝑥∑ ∑ ∑ 𝑎𝑖,𝑗,𝑚.𝑘 ∈𝑖,𝑗,𝑘𝑗𝑖 (5)
𝑎𝑖,𝑗,𝑚 denotes the weight coefficients contribution to each measurement point with indexes i, j for
ring number “m” (Figure 3).
Figure 3: Schematic representation of the wire mesh analysis for azimuthal averaging of gas
hold up profiles.
19
4. Results and discussion
4.1 Overall volume averaged hold-up
The overall gas fraction was estimated using the pressure drop method and expanded liquid
height method. Experiments were performed for the two different spargers and seven gas
velocities. The gas fraction profiles obtained from these methods are reported in Figure 4. Figure
4A reports the overall grass fraction, for both point sparger and four arm sparger, measured from
the pressure drop method. The top and bottom most pressure transducers were considered in these
calculations in order to measure the overall gas fraction, Eq. 7 was used for calculating the gas
fraction from the pressure drop method. For both the sparger designs, the gas fraction value
increases with an increase in the gas velocity. The gas fraction observed for the four arm sparger
is consistently higher than the point sparger. Figure 4B reports the gas fraction values measured
using the expanded height measurements. Prior to each experiment, the initial liquid height was
measured and after the gas injection at a specified flow rate, the expanded liquid velocity is
measured. The measured values were substituted in the Eq. 1 to calculate the volume averaged gas
hold-up. The change in gas fraction with the gas velocities for two different spargers indicates that
the void fraction increases with increase in gas velocities and the four arm sparger offer higher gas
hold-up when compared to the point arm sparger. From the void fraction measurement methods,
it is clear that, for the two sparger designs, the two-phase flow is changing from homogenous to
heterogeneous flow regime at a superficial gas velocity of 0.15 m/s.
20
Figure 4A: Overall gas fraction measured from PTs data, for the initial third water level.
21
Figure 4B: Overall gas holdup for the third water level measured from expanded liquid heights
data
22
4.2 Effect of Gas velocity
4.2.1 Near the gas sparger
The effect of gas velocity on the void fraction profiles is presented in the following section.
Two different representations of gas fraction values were presented in this study, one
representation is surface plots and the other representation is radial profiles. Seven different gas
velocities (0.03 to 0.26 m/s) were maintained to cover both homogenous and heterogeneous flow
regimes. Figure 5A shows the surface plots of the gas fraction values recorded using wire mesh
plane located at 10 cm away from gas sparger. The surface plots indicate that the four peak gas
fraction values are observed exactly above the four-arm gas sparger. Which can be attributed to
the gas jet coming out of the gas sparging arms results into a maximum gas fraction.
Figure 5A: Surface plots of gas fraction near the four arm sparger (wire mesh location 1) of
different fluid concentrations for the superficial gas velocity of 0.074m/s
23
Further, with an increase in the gas velocity, the maximum gas fraction also increases. The radially
averaged gas fraction profiles were calculated by averaging the surface plots over 80 azimuthal
rings. Figure 5B reports the radial gas fraction profiles for different fluids at a constant superficial
gas velocity of 0.074 m/s. The radial gas fraction profile is quite interesting and not parabolic in
nature. The gas fraction value near the center of the column and near the column wall. As we move
away from the center of the column gas fraction increases from a low value at the center and
increases to a maximum value at around 40 mm away from the center. Moving further away from
the center the gas fraction values gradually decrease till 140 mm away from the center and drops
to a minimum near the column wall. The same observation was reported for all of the gas velocities
investigated in the present work (Please refer to Figure 7).
Figure 5B: Radial profiles of gas fraction near the four arm sparger(wire mesh location 1) of
different fluid concentrations for the superficial gas velocity of 0.074m/s
24
The maximum gas fraction near the sparger region was high for the CaCl2 solution when compared
to the other fluids studied in this work. Though there is a difference in the magnitude of the gas
fraction values, the nature of the gas fraction profiles is nearly the same for all the fluids.
Figure 6A: Surface plots of gas fraction near the four arm sparger (wire mesh location 1) of
different fluid concentrations for the superficial gas velocity of 0.221 m/s
Figure 6A reports the surface plots gas fraction for the superficial gas velocity of 0.221 m/s. and
Figure 6B reports the radial gas fraction profiles obtained by averaging the over 80 azimuthal
rings. The surface plots and radial gas fraction profiles are exactly identical to those observed for
the low superficial gas velocity (refer Figure 5) except the magnitude of the gas fraction values.
25
As evident from both the figures 5 and 6, the maximum gas fraction values are 0.40-0.51 for the
gas velocity of 0.074 m/s, where it ranges from 0.7 to 0.74 for the superficial gas velocity of 0.221
m/s.
Figure 6B: Radial profiles of gas fraction near the four arm sparger (wire mesh location 1) of
different fluid concentrations for the superficial gas velocity of 0.221 m/s
Figure 7 reports the change in the gas holdup profiles for air-tap water system near the four-arm
gas sparger for all the superficial gas velocities. Though the nature of the gas fraction profiles is
the same for all the gas velocities the gas fraction values consistently increase with an increase in
the gas velocity. For low gas velocity, the gas fraction values were not varying with the radial
26
location. But for the high gas velocity (0.26 m/s), the gas fraction profiles are showing two peaks.
In other words, the non-uniformity of the radial profiles increases with an increase in the gas
velocity.
Figure 7: Radial profiles of gas fraction near the four arm sparger (wire mesh location 1) of Tap
Water for all the superficial gas velocities (0.0347 to 0.258 m/s).
4.2.2 In the fully developed region
Similar to the results reported in the previous section 4.2.1, this section presents similar
plots for gas fraction values in the fully developed region. The wire mesh sensor was installed
away from the sparger so that the gas fraction profiles are not affected by the gas entering and
exiting the bubble column reactor. The figure reports both surface plot and radial gas fraction
27
profiles for four gas-liquid systems (Tap water, DI water, Na2CO3, CaCl2). Unlike the profiles near
the gas sparger, the hold-up profiles and surface plots in the fully developed region are parabolic
in nature. The maximum gas fraction values are in the range of 0.26-0.31 for the superficial gas
velocity of 0.074 m/s (Figure 8).
Figure 8A: Surface plots of gas fraction in the fully developed region (wire mesh location 3) of
the different fluid concentrations for the superficial gas velocity of 0.074m/s.
28
Figure 8B: Radial profiles of gas fraction in the fully developed region (wire mesh location 3) of
the different fluid concentrations for the superficial gas velocity of 0.074 m/s
The gas fraction profiles in figure 8b show that the gas hold up is high in the center of the column
and reduces gradually near the column wall. The change in the gas fraction over the radial direction
results into density gradient causing the liquid circulations when the bubble column operated in
the batch mode.
29
Figure 9A: Surface plots of the gas fraction in the fully developed region (wire mesh location 3)
of the different fluid concentrations for the superficial gas velocity of 0.221m/s.
30
Figure 9B: Radial profiles of the gas fraction in the fully developed region (wire mesh location
3), for the superficial gas velocity of 0.221 m/s
Figure 9 reports the gas fraction data in the form of the surface plots and radial void fraction
profiles for the four different fluid systems at a superficial gas velocity of 0.221 m/s. The range of
the void fraction is varied from 0.46 to 0.51 for all different fluids at a superficial gas velocity of
0.221 m/s. The gas fraction profiles are parabolic in nature and the gas fraction was high near the
column center and low near the column wall. The higher the gas hold-up in the center means that
the liquid is being carried away by the gas in the central region and comes down near the column
wall.
31
Figure 10: Radial profiles of the gas fraction in the fully developed region (wire mesh location
3) for all the superficial gas velocities (0.0347 to 0.258 m/s).
Figure 10 reports the radial void fraction profiles for all the seven superficial gas velocities studied
in this work (0.037 to 0.258 m/sec). For the lowest superficial gas velocity, the gas hold-up profile
was almost flat, and the steepness of the gas hold-up profiles are becoming steep as the gas velocity
is increasing. Interestingly, for all the gas velocities, the gas hold-up profiles were almost flat from
the column center to the radial distance of 40 mm. Beyond the 40 mm radial distance the gas
fraction profile are gradually reducing to the lower values near the column wall.
32
4.3 Effect of initial liquid height
The hydrostatic head over the gas sparger has a significant effect on the gas fraction values
and profiles. In the present work, an attempt has been made to understand the effect of the
hydrostatic head by varying the initial liquid heights. Three initial liquid heights were maintained,
for first water level the initial liquid level was maintained at 1.1 m, 1.5 m for second water level
and 1.9 m from the bottom of reactor for the third water level. Figure 11 reports the effect of the
superficial gas velocity on the gas hold-up for different initial liquid velocities and different
sparger designs. The gas fraction was estimated using the expanded liquid heights methods (using
Eq. 1).
Figure 11A shows the change in gas hold-up with superficial gas velocity for the two
different spargers for the first water level. The maximum gas hold-up was recorded for this water
level is different for both the spargers and four arm sparger offers high gas hold-up compared to
the point sparger. The difference in these void fractions increases with an increase in gas velocity.
For the point sparger, the maximum gas fraction is 0.23 for the superficial gas velocity of 0.26
m/s, whereas, for the point sparger the maximum gas fraction is 0.33 for the same gas velocity of
0.26 m/s. For both the spargers, the transition from the homogenous flow regime to the
heterogeneous flow regime occurs at a gas velocity of 0.12 m/s.
33
Figure 11A: Overall void fraction measured from expanded liquid heights for the first water level
Figure 11B reports the change in void fraction with superficial gas velocity for both
spargers at the second water level. The maximum gas fraction recorded for this water level is 0.32
for the point sparger and 0.42 for the four arm sparger at a superficial gas velocity of 0.26 m/s.
The transition from the homogenous flow regime to a heterogeneous flow regime, (it is only
observed for point sparger) occurs at 0.12 m/s which is the same as for the first water level. Even
for this initial water level, the difference in gas fraction values for both spargers increases with an
increase in gas velocity.
34
Figure 11B: Overall void fraction measured from expanded liquid heights, for the second water
level.
35
Figure 11C: Overall void fraction measured from expanded liquid heights, for the third water
level
Figure 11C shows the change in void fraction with superficial gas velocity for the third
water level. For the four arm sparger, the maximum gas fraction is found to be 0.5 for the
superficial gas velocity of 0.26 m/s, whereas for the point sparger, the maximum gas hold-up was
recorded for this water level is 0.42 for the superficial gas velocity of 0.26 m/s. Due to the
hydrostatic head is high for the third water level the gas fraction, the difference in void fraction
profiles from both the spargers are very close. As the initial water level increases the gas injected
36
into the reactor experiences very high resistance before exiting the reactor results in higher gas
phase residence time and hence the higher hold-up.
4.4. Effect of sparger design
As it is very known that the gas sparger design plays an important role in the performance
of the bubble column reactors. Two different spargers: (a) point sparger, (b) four-arm sparger were
used in the present study. For the point sparger, the gas was injected with a straight pipe with an
ID of 25.4 mm. For the four arm sparger, the gas was injected using “cross” shaped four arms. The
gas injected from each arm with equally spaced holes on each arm. As mentioned before the wire
mesh sensors were placed strategically along the column height to capture the flow profile near
the sparger and in the fully developed region.
Figure 12(A) Radial void fraction profiles in the sparger region for all superficial gas velocities
and the gas was sparged using point sparger. (B) Radial void fraction profiles in the sparger region
for all superficial gas velocities and the gas was sparged using four arm sparger.
Figure 12 reports the radially averaged void fraction profiles for both point and four arm
spargers, the data was obtained from the wire mesh plane that is located near the sparger. The
37
experiments were performed for the air-tap water system and the gas velocity was varied from the
0.037 to 0.26 m/s. As it is evident from the above figures the gas fraction profiles were completely
different for both spargers. For the point sparger (Figure 12A), the void fraction profiles for the
different superficial gas velocities used indicate a conical shape with the highest gas fraction in the
center of the column and further becomes almost zero near the radial distance of 80 mm from the
center of the column. Whereas for the four arm sparger, the gas fraction profiles are quite different.
The gas fraction values are low near the column wall and column center.
Figure 13 (A) Radial void fraction profiles in the fully developed region for all superficial gas
velocities and the gas was sparged using point sparger. (B) Radial void fraction profiles in the fully
developed region for all superficial gas velocities and the gas was sparged using four arm sparger.
Figure 13 shows the radial void fraction profiles for the air-tap water system for all seven gas
velocities investigated and two spargers used in the present study. Though the gas was sparged
using two different sparges, the radial gas hold-up profiles are parabolic in nature in the fully
developed region. The gas injected from both the spargers in the form of bubbles as they rise
upwards the bubbles will coalesce and the effect of the sparger was nullified as the bubbles move
38
away from the sparger. In order to exactly find the axial location where the sparger effect is fading
off, more experiments are needed with more axial locations of wire mesh sensors.
5. Conclusions
In the present study, wire mesh sensors and pressure transducers were successfully
employed to investigate void fraction measurements in the large diameter batch bubble columns.
The working principle of wire mesh sensor and the analysis of the wire mesh data for explained in
detail. Experiments were performed for seven superficial gas velocities over a wide range (0.037
to 0.26 m/s) to accommodate both the homogenous and heterogeneous flow regimes. Two different
spargers (point and four arm spargers) were used to study the effect of the sparger on the
hydrodynamics. The effect of superficial gas velocity and sparger design on the radial and steady-
state surface plots for void fraction have been reported. For a given superficial gas velocity, the
four arm sparger was giving higher gas fraction when compared to the point sparger. Regime
transition was clearly observed for point sparger and it was less observed for the four arm sparger.
For the fluids under consideration, the Air-CaCl2 aqueous solution offers a higher void fraction
than the other three air-water systems. Whereas, this trend is not observed in the fully developed
region. It is also found that the gas sparger has an effect on the gas fraction profile near the sparger
region whereas the sparger design has no effect on the void fraction profile in the fully developed
region.
39
6. An additional chapter on the Very-High-Temperature Reactor (DOE Project) and the
Drilling Fluid (PIRE Project):
The Very High-Temperature Reactor is one of the 6 technologies classified by the
Generation IV International forum as a promising reactor type likely to power our world in the
coming decades. These designs ensure passive cooling in case of loss of forced circulation or
pressure. To this end, a high-temperature gas flow loop, operable with any gas (air, CO2, N2, He,
etc.) at high pressure and high temperature (up to 7 MPa at 760oC inlet gas temperature) has been
constructed at City College of New York. In this project, the forced and natural convection heat
transfer characteristics of air, nitrogen, and helium at high pressure and high-temperature
conditions are being investigated. The main objective of this project is to obtain reliable thermal-
hydraulic data for the development and validation of VHTR design and safety analysis codes in
coordination with other DOE-supported projects. (Francisco I. Valentin et al, 2014). Test facility
improvements like the installation of a double stage reciprocating (1500 PSI) compressor, addition
of thermocouples and adjustment of the loop piping system were done on the existing set-up, after
which experiments were performed to analyze the mass flow rate of two gases (Helium and
Nitrogen) at different concentrations which were circulating in two graphite body channels inside
the reactor. A time span of about 8 hours was chosen for each experiment and the operating
temperature was selected to be between 100oC and 400oC, with a 100oC increment, to examine the
transportation of the gases from the bottom plenum to the upper plenum within the reactor.
The settling of barite in Drilling Fluid (PIRE Project) was studied initially with an X-ray
source and its camera to analyze the settling of the fluid composition in different test section sizes,
(ranging from 2mm, 5mm, and 1cm) at steady state. Further study was done using a Gamma
densitometry setup, with the fluid also placed in varying test section dimensions to analyze the
40
behavior of the settling of barite in steady state. Upon an extended study of the settling of barite in
drilling fluid, using both the experimental methods, it was observed that the drilling muds forms a
gel like structure which results in the suspension of barite particles over a long period of time. The
fluid did not present a normal settling of barite in static cell meaning without any shear.
Experiments are being currently carried out to study the settling behavior of barite, a major
component of the fluid, in water. Also, a model of the drilling process which introduces shear into
the system is also being inculcated in the current study to further understand and analyze the
drilling fluid.
41
References:
Buwa, V. V. and V. V. Ranade, “Dynamics of Gas-Liquid Flow in a Rectangular Bubble Column:
Experiments and Single/Multi-Group CFD Simulations,” Chem. Engng. Sci., 57, 4715-4736
(2002).
Chen, R. C., and L. S. Fan, “Particle Image Velocimetry for Characterizing the Flow Structure in
Three-Dimensional Gas-Liquid-Solid Fluidized Beds,” Chem. Engng. Sci., 47, 3615 (1992)
Degaleesan S., Turbulence and Liquid Mixing in Bubble Columns. D.Sc. Thesis, Washington
University, USA, 1997.
Devanathan, N., “Investigation of Liquid Hydrodynamics in Bubble Columns via Computer
Automated Radioactive Particle Tracking.” D.Sc. Thesis, Washington University, USA
(1991).
Fan, L. S., “Gas-Liquid-Solid Fluidization Engineering.” Butterworth-Heinemann, Boston (1989)
Hernandez-Alvarado, F., Kalaga D.V, Turney D.E, Banerjee S, Joshi J.B, Kawaji M. 2017. Void
fraction, bubble size and interfacial area measurements in co-current down-flow bubble
column reactor with microbubble dispersion. Chemical Engineering Science. 168 (31), 403–
413.
Kalaga, D. V., Pant, H. J., Dalvi, S. V., Joshi, J. B. and Roy, S. (2017), Investigation of
hydrodynamics in a bubble column with internals using radioactive particle tracking (RPT).
AIChE J., 63: 4881–4894. doi:10.1002/aic.15829.
Krishna, R., J. W. A. de Swart, D. E. Hennephof, J. Ellenberger and H. C. J. Hoefsloot, “Influence
of Increased Gas Density on Hydrodynamics of Bubble Column Reactors,” AIChE J., 40, 112-
119 (1994).
Krishna, R., P. M. Wilkinson, and L. L. van Dierendonck, “A Model for Gas Holdup in Bubble
Columns Incorporating the Influence of Gas Density on Flow Regime Transitions,” Chem.
Engng. Sci., 46, 2491-2496 (1991)
L.C. Mutharasu, Kalaga, D.V., Mayur Sathe, Damon E. Turney, Derek Griffin, Xueliang Li,
Masahiro Kawaji, Krishnaswamy Nandakumar, J.B. Joshi, Experimental study and CFD
simulation of the multiphase flow conditions encountered in a Novel Down-flow Bubble
Colum, Chemical Engineering Journal, 2018.
Pfleger, D., S. Gomes, N. Gilbert, and H.G. Wagner, “Hydrodynamic Simulations of Laboratory
Scale Bubble Columns Fundamental Studies of the Eulerian-Eulerian Modeling Approach,”
Chem. Engng. Sci., 54, 5091-5099 (1999).
42
Prasser, Böttger, A., Zschau, J., 1998. A new electrode-mesh tomograph for gas-liquid flows. Flow
Meas. Instrum. 9, 111–119.
Sanyal, J., S. Vasquez, S. Roy, and M. P. Dudukovic, “Numerical Simulation of GasLiquid
Dynamics in Cylindrical Bubble Column,” Chem. Engng. Sci., 54, 5071-5083 (1999)
Shah, Y. T., B. G. Kelkar, S. P. Godbole, and W. D. Deckwer, “Design Parameter Estimations for
Bubble Column Reactors,” AIChE. J., 28, 353-379 (1982).
Silva, M.Da, 2008. Impedance sensors for fast multiphase flow measurement and imaging.
Sokolichin, A. and G. Eigenberger, “Gas-Liquid Flow in Bubble Columns and Loop Reactors: Part
I. Detailed Modelling and Numerical Simulation,” Chem. Eng. Sci., 49, 5735 (1994).
Tatterson, G. B., “Fluid Mixing and Gas Dispersion in Agitated Tanks.” McGraw-Hill (1991).